1 Biomark, Inc.
The Bureau of Reclamation (BOR), Idaho Governor’s Office of Species Conservation (OSC), and an interdisciplinary team of partners have assembled an Upper Salmon Assessment Team to complete biologic and geomorphic analyses to support future project identification, prioritization, and inform restoration design in the Upper Salmon Subbasin, Idaho. The biologic and geomorphic analyses are being lead by Biomark Inc. (Biomark) and Rio Applied Science and Engineering (Rio ASE), respectively. Past efforts from the team resulted in the development of a watershed-scale Integrated Rehabilitation Assessment (IRA; Idaho OSC Team (2019)) in the Lemhi, Pahsimeroi, and Upper Salmon (Sawtooth Valley) watersheds. This initial phase of the project identified, generally, the “problem” by spatially quantifying capacity limitations for spring/summer Chinook salmon and summer run steelhead within a geomorphic context across these three watersheds. The second phase, termed the Multiple Reach Assessments (MRA), includes identifying appropriate and focused “solutions” to the acknowledged capacity problems within four valley segments: Upper Lemhi, Lower Lemhi, Lower Pahsimeroi, and Upper Salmon (Decker Flats). To achieve this goal, the team will collaboratively summarize existing and targeted physical habitat conditions relative to documented habitat needs for specific species and life stages, including discussion of high-quality habitat creation and maintenance to inform future rehabilitation actions.
In the IRA, it was determined that, for Chinook salmon, habitat capacity was limited to support presmolts during winter months, and to a slightly lesser degree, parr during summer months. Habitat was not found to be limiting for adult spawning. For steelhead, habitat capacity was identified as limiting for juvenile rearing, at least in the Pahsimeroi River; again, habitat was not found to be limiting for adult spawning. The available habitat capacity was estimated using quantile regression forest (QRF; IRA Appendix B) whereas habitat requirements were estimated for current escapement and recovery goals using a generalized capacity model (IRA Appendix C).
The goal of this document is to further assess existing conditions and evaluate the hydraulic suitability, particularly depth and velocity, of the four target valley segments to support select life stages of Chinook salmon and steelhead. By comparing depth and velocity suitability curves for Chinook salmon and steelhead, developed specifically for the Salmon River watershed (Maret et al. 2006), to continuous modeled depths and velocities (supported by bathymetric Light Detection and Ranging; LiDAR) available for the four valley segments, we can further our understanding of how habitat, related specifically to hydraulics and the governing morphology, may be limiting recovery of Chinook salmon and steelhead in the Upper Salmon subbasin. This information can help identify geomorphic reaches where existing depth and velocity may be limiting particular species and life stages, which could prove useful for project prioritization.The incorporation of a multitude of data sources (QRF, publication review, morphological analysis, hydraulic habitat suitability, etc.) allows for a robust assessment of the habitat and limiting factors for target species and life stages.
Evaluate the composite suitability of geomorphic reaches in the upper Lemhi, lower Lemhi, lower Pahsimeroi, and upper Salmon (Decker Flat) valley segments based on modeled depth and velocity rasters supported by a multitude of LiDAR sampling events and sensors that incompass the study reaches. Composite suitability is evaluated for both Chinook salmon and steelhead at multiple life stages including, adult spawning and juvenile rearing at various discharge scenarios (see Table 2.2). The proportion of each geomorphic reach classified as simple, mixed, or complex is also provided for reference.
We provide methods for calculating the composite (depth & velocity) suitability, at each pixel, for a single scenario. These pixels are then summarized by river kilometer (rkm) and geomorphic reach, including visualizations of the results. Which river kilometers are contained within each geomorphic reach is shown in Table 2.1. A given scenario includes a watershed, species, life stage, and season combination. Here, we provide detailed methods for Scenario 1 in Table 2.2: Lemhi River, Chinook salmon, juvenile summer (parr) rearing. The same methods were then applied across all scenarios utilizing the appropriate depth and velocity suitability curves (depending on species and adult versus juvenile) or differing input depth and velocity rasters (depending on season and estimated discharge). All scenarios evaluated are summarized in Table 2.2. All data, scripts, outputs, and reports for this analysis can be found within the mra_hsi repository at https://github.com/mackerman44/mra_hsi.
Detailed methods are as follows:
Raster .tifs containing depth and velocity values were imported into R (R Core Team 2019). For the Lemhi River, raster pixels were 1m x 1m. As an example, for the Lemhi River, summer, low-flow scenario rasters were named D_Aug_All.tif and V_Aug_All.tif. Rasters were obtained from a previously funded study in the Lemhi River (Tonina et al. 2019), where appropriate discharge volumes for hydraulic analysis were analyzed based on the Lemhi River Base Model (Borden 2016).
Import polygon shapefiles delineating the river kilometers and geomorphic reaches as defined in the IRA. These shapefiles are used to interrogate the depth and velocity .tifs to determine which rkm or geomorphic reach each depth and velocity pixel falls within.
Read in the depth and velocity habitat suitability curves for Chinook salmon and steelhead from Maret et al. (2006). Depth and velocity habitat suitability index (HSI) curves were available for both species for the adult spawning and juvenile rearing life stages. Functions to calculate the suitability for a given depth or velocity are available in the mra_hsi repository in the R/ directory. The Habitat Suitability section below shows the HSI curves used from Maret et al. (2006).
Use the HSI curves to calculate the depth and velocity suitability for each raster pixel. The result is two new rasters each containing the calculated depth and velocity suitability values, respectively.
Calculate the composite suitability value for each raster pixel as the geometric mean of the depth and velocity suitability values. The result is a third composite suitability raster.
Extract the composite suitability values located within each rkm for each pixel into an R dataframe. The dataframe can then be used to summarize and visualize the composite suitability by rkm or geomorphic reach for the given watershed, species, life stage, and discharge scenario.
Write out the composite suitability values, rkms and geomorphic reaches for each pixel to a .csv. These results are all stored in the output/hsi_raw/ and output/hsi_rkm/ directories in the mra_hsi repository.
Calculate the total wetted area, weighted usable area, and normalized weighted usable area (i.e. hydraulic habitat suitability) for each species, life stage, and rkm / geomorphic reach. The total wetted area was calculated by counting the total number of pixels with a known area, that occur within each rkm and wetted area for a given scenario. The weighted usable area (WUA) was calculated by summing the composite suitability values of all pixels within that same area. And finally, the normalized weighted usable area i.e. hydraulic habitat suitability (HHS), was calculated by dividing the WUA by the total wetted area for each rkm or geomorphic reach.
Finally, we visualized the composite suitability values by valley segment, species, life stage (adult spawning, juvenile summer rearing, juvenile winter rearing), and rkm / geomorphic reach. Violin plots showing the distribution of composite suitability values are provided in the Results section. For reference, the proportion of each reach classified as simple, mixed, or complex is also provided. Further, we provide maps showing the hydraulic habitat suitability values by species and life stage for each valley segment, and plots illustrating the weighted usable area and normalized weighted usable area for each individual scenario.
The resulting raster .tifs showing depth, velocity, and composite suitability are too large to store in a https://github.com/ repository, but are available from the authors upon request.
| River | GeoReach | min | max |
|---|---|---|---|
| Lemhi River | GR_1 | 0 | 3 |
| Lemhi River | GR_2 | 4 | 12 |
| Lemhi River | GR_3 | 13 | 16 |
| Lemhi River | GR_4 | 17 | 21 |
| Lemhi River | GR_5 | 22 | 25 |
| Lemhi River | GR_6 | 26 | 28 |
| Lemhi River | GR_7 | 29 | 37 |
| Lemhi River | GR_8 | 38 | 40 |
| Lemhi River | GR_9 | 41 | 44 |
| Lemhi River | GR_10 | 45 | 47 |
| Lemhi River | GR_11 | 48 | 52 |
| Lemhi River | GR_12 | 53 | 61 |
| Lemhi River | GR_13 | 62 | 71 |
| Lemhi River | GR_14 | 72 | 79 |
| Lemhi River | GR_15 | 80 | 86 |
| Lemhi River | GR_16 | 87 | 93 |
| Pahsimeroi River | GR_8 | 48 | 52 |
| Pahsimeroi River | GR_9 | 53 | 69 |
| Pahsimeroi River | GR_10 | 70 | 86 |
| Salmon River | GR_4 | 11 | 16 |
| Salmon River | GR_5 | 17 | 21 |
| Salmon River | GR_6 | 22 | 31 |
| Salmon River | GR_7 | 32 | 35 |
| Salmon River | GR_8 | 36 | 44 |
| Salmon River | GR_9 | 45 | 55 |
| Salmon River | GR_10 | 56 | 56 |
Figures 2.1 and 2.2 show HSI curves for Chinook salmon and steelhead from Maret et al. (2006), respectively, used the analysis. These curves are used to calculate the depth or velocity suitability value for each pixel within a scenario. The composite suitability for a pixel is then calculated as the geometric mean of those values. As an example, using juvenile Chinook salmon and depth, if a pixel has a depth of 0m, that pixel is assigned a suitability of 0 whereas if the depth is greater than approximately 0.6m it is assigned a suitability of 1; a depth of 0.5m would be assigned a suitability of \(\sim\) 0.7.
Figure 2.1: Suitability indices at varying depths and velocities for juvenile rearing and adult spawning for Chinook salmon from Maret et al. (2006).
Figure 2.2: Suitability indices at varying depths and velocities for juvenile rearing and adult spawning for steelhead from Maret et al. (2006).
| Scenario | Watershed | Species | Life Stage | Season | Depth Raster | Velocity Raster |
|---|---|---|---|---|---|---|
| 1 | Lemhi | Chinook | Juvenile | Summer | D_Aug_All.tif | V_Aug_All.tif |
| 2 | Lemhi | Chinook | Juvenile | Winter | d_jan_v2.tif | v_jan_v2.tif |
| 3 | Lemhi | Chinook | Spawning | Summer | D_Aug_All.tif | V_Aug_All.tif |
| 4 | Lemhi | Steelhead | Juvenile | Summer | D_Aug_All.tif | V_Aug_All.tif |
| 5 | Lemhi | Steelhead | Juvenile | Winter | d_jan_v2.tif | v_jan_v2.tif |
| 6 | Lemhi | Steelhead | Spawning | Spring | d_jan_v2.tif | v_jan_v2.tif |
| 7 | Pahsimeroi | Chinook | Juvenile | Summer | - | - |
| 8 | Pahsimeroi | Chinook | Juvenile | Winter | Pah_WLow_depth.tif | Pah_WLow_velocity.tif |
| 9 | Pahsimeroi | Chinook | Juvenile | Spring | Pah_1pt5_depth.tif | Pah_1pt5_velocity.tif |
| 10 | Pahsimeroi | Chinook | Spawning | Summer | Pah_WLow_depth.tif | Pah_WLow_velocity.tif |
| 11 | Pahsimeroi | Steelhead | Juvenile | Summer | - | - |
| 12 | Pahsimeroi | Steelhead | Juvenile | Winter | Pah_WLow_depth.tif | Pah_WLow_velocity.tif |
| 13 | Pahsimeroi | Steelhead | Juvenile | Spring | Pah_1pt5_depth.tif | Pah_1pt5_velocity.tif |
| 14 | Pahsimeroi | Steelhead | Spawning | Spring | Pah_1pt5_depth.tif | Pah_1pt5_velocity.tif |
| 15 | Upper Salmon | Chinook | Juvenile | Summer | US_Summer75_depth.tif | US_Summer75_velocity.tif |
| 16 | Upper Salmon | Chinook | Juvenile | Winter | US_Winter75_depth.tif | US_Winter75_velocity.tif |
| 17 | Upper Salmon | Chinook | Juvenile | Spring | US_1pt5year_depth.tif | US_1pt5year_velocity.tif |
| 18 | Upper Salmon | Chinook | Spawning | Summer | US_Summer75_depth.tif | US_Summer75_velocity.tif |
| 19 | Upper Salmon | Steelhead | Juvenile | Summer | US_Summer75_depth.tif | US_Summer75_velocity.tif |
| 20 | Upper Salmon | Steelhead | Juvenile | Winter | US_Winter75_depth.tif | US_Winter75_velocity.tif |
| 21 | Upper Salmon | Steelhead | Juvenile | Spring | US_1pt5year_depth.tif | US_1pt5year_velocity.tif |
| 22 | Upper Salmon | Steelhead | Spawning | Spring | US_1pt5year_depth.tif | US_1pt5year_velocity.tif |
We evaluated 22 scenarios in total which are summarized in Table 2.2. For all summer scenarios, we used rasters from a discharge scenario representative of low flow conditions. In the case of the Lemhi and Pahsimeroi watersheds, all spring and winter scenarios used rasters from moderate to high flow conditions. Alternatively, for the Upper Salmon, winter scenarios were analyzed using a low-flow discharge. Finally, for the Upper Salmon, we added spring scenarios for juvenile rearing to evaluate high-flow conditions, as both summer and winter are typically low-flow in that watershed.
Here, we provide a summary of the distribution (as violin plots) and mean (as maps) of composite suitability values by valley segment, species, life stage, and geomorphic reach. Raw outputs and raster .tifs of depth, velocity, and composite suitability values are available in the mra_hsi repository or from the authors.
Figure 3.1 summarizes the composite hydraulic suitability within the Upper Lemhi valley segment to support Chinook salmon spawning and juvenile rearing (summer and winter) along with the proportion of each geomorphic reach classified as simple, mixed, or complex. In general, the hydraulic suitability for spawning, for both Chinook salmon and steelhead, is high, with a large number of pixels having a suitability of 1. Hydraulic suitability for steelhead juvenile rearing, during both the summer and winter scenarios, also tends to be high in the Upper Lemhi valley segment; whereas suitability for juvenile Chinook salmon rearing is low. Interestingly, geomorphic reach 01, the upstream-most reach starting at Leadore, contains more pixels with a suitability above 0 than all other geomorphic reaches for both summer and winter juvenile rearing.
Figure 3.1: Violin plots showing the distribution of composite suitability values (geometric mean of depth and velocity suitability) across geomorphic reaches in the Upper Lemhi valley segment. Results for both Chinook salmon and steelhead and for three lifestages (adult spawning, juvenile summer rearing, juvenile winter rearing) are shown. The bottom panel shows the proportion of each geometric reach classified as simple, mixed, or complex.
Figure 3.2 shows the hydraulic habitat suitability of each river kilometer by species and life stage, with the outlines of the geomorphic reaches in the Upper Lemhi valley segment. The results parallel those of Figure 3.1. Hydraulic suitability for spawning (both species) and for steelhead rearing tends to be high (red) whereas suitability for Chinook salmon rearing tend to be low (blue).
Figure 3.2: Map showing the hydraulic habitat suitability by life stage for Chinook salmon and steelhead in the Upper Lemhi valley segment, plotted by river kilometer with geomorphic reaches also shown.
Figure 3.3: The weighted usable area (WUA) shown as points and on the primary axis and the hydraulic habitat suitability (HHS) i.e. normalized WUA shown as bars and on the secondary axis by species, life stage, and river kilometer for the upper Lemhi River valley segment. The HHS is normalized by dividing the weighted usable area for each reach by the total area of that reach.
Figure 3.4 summarizes the composite hydraulic suitability within the Lower Lemhi valley segment to support Chinook salmon spawning and juvenile rearing (summer and winter) along with the proportion of each geomorphic reach classified as simple, mixed, or complex. Similar to the Upper Lemhi, hydraulic suitability for spawning in the Lower Lemhi is high; likewise, suitability for steelhead juvenile rearing is high (i.e., a large number of pixels have a composite suitability approaching or eual to 1). Suitability for juvenile Chinook salmon rearing in the Lower Lemhi is low. Coincidently, reaches with a high proportion classified as simple (e.g., GR_09-11) have a very high proportion of pixels with a suitability at 0; whereas the most complex reach, GR_15, has a much larger proportion of pixels landing higher in the violin distribution.
Figure 3.4: Violin plots showing the distribution of composite suitability values (geometric mean of depth and velocity suitability) across geomorphic reaches in the Upper Lemhi valley segment. Results for both Chinook salmon and steelhead and for three lifestages (adult spawning, juvenile summer rearing, juvenile winter rearing) are shown. The bottom panel shows the proportion of each geometric reach classified as simple, mixed, or complex.
Figure 3.5 shows the hydraulic habitat suitability of each river kilometer by species and life stage, with the outlines of the geomorphic reaches in the Lower Lemhi valley segment. The Lower Lemhi map shows the same trends as the Upper Lemhi map. Hydraulic suitability for spawning and juvenile steelhead rearing is adequate; suitability for juvenile Chinook rearing is poor. Note that the upstream simple reaches (GR_09-11) are the deepest blue whereas the ‘more’ complex GR_15 seems to support higher calculated suitability (lighter blue) for juvenile Chinook rearing.
Figure 3.5: Map showing the hydraulic habitat suitability by life stage for Chinook salmon and steelhead in the Lower Lemhi valley segment, plotted by river kilometer with geomorphic reaches also shown.